Re: [Xen-devel] [PATCH] xen-blkfront: fix mq start/stop race

[Junxiao Bi <junxiao.bi@xxxxxxxxxx>]

Hi Boris & Juergen,

Could you help review this patch? This is a race and will cause the io hung.

Thanks,
Junxiao.

On 06/22/2017 09:36 AM, Junxiao Bi wrote:
> When ring buf full, hw queue will be stopped. While blkif interrupt consume
> request and make free space in ring buf, hw queue will be started again.
> But since start queue is protected by spin lock while stop not, that will
> cause a race.
> 
> interrupt:                                      process:
> blkif_interrupt()                               blkif_queue_rq()
>  kick_pending_request_queues_locked()
>   blk_mq_start_stopped_hw_queues()
>    clear_bit(BLK_MQ_S_STOPPED, &hctx->state)
>                                                  blk_mq_stop_hw_queue(hctx)
>    blk_mq_run_hw_queue(hctx, async)
> 
> If ring buf is made empty in this case, interrupt will never come, then the
> hw queue will be stopped forever, all processes waiting for the pending io
> in the queue will hung.
> 
> Signed-off-by: Junxiao Bi <junxiao.bi@xxxxxxxxxx>
> Reviewed-by: Ankur Arora <ankur.a.arora@xxxxxxxxxx>
> ---
>  drivers/block/xen-blkfront.c |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
> index 8bb160cd00e1..4767b82b2cf6 100644
> --- a/drivers/block/xen-blkfront.c
> +++ b/drivers/block/xen-blkfront.c
> @@ -912,8 +912,8 @@ out_err:
>       return BLK_MQ_RQ_QUEUE_ERROR;
>  
>  out_busy:
> -     spin_unlock_irqrestore(&rinfo->ring_lock, flags);
>       blk_mq_stop_hw_queue(hctx);
> +     spin_unlock_irqrestore(&rinfo->ring_lock, flags);
>       return BLK_MQ_RQ_QUEUE_BUSY;
>  }
>  
> 


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel